Showing posts with label Image Recognition. Show all posts
Showing posts with label Image Recognition. Show all posts

Saturday, July 12, 2014

Layar-ing Blippar - Post-acquisition thoughts on major shake-up in AR browser market



BARCELONA - If you’re into AR and have played around with the many AR-enabled mobile apps published in iTunes and Google Play, you have probably heard by now about Blippar acquiring Layar. This means different things for everyone involved in the AR industry, for consumers, for professionals and for the future of the AR industry as a whole.

I first came into contact with Layar at a startup event in Barcelona in 2009 and was impressed with Layar's vision and ability to market AR's potential for immersive experiences on mobile to both techies and marketeers alike. I also described Layar's use of geo-location in my book on Location Aware Applications.

If you’re not entirely certain what those companies do, the rundown is that they both provide an app which enables consumers to scan printed media or packaged products that have an interactive campaign or digital content attached to them. Typically users know about that content because there is a call to action. Those apps compete with others like Junaio or Wikitude, alternative ‘AR browsers’.

The most prominent advantage of AR browsers is that users need only one app for multiple content. Once installed, it pulls new content on demand from the cloud.The disadvantage when comparing branded apps and browsers is for content creators. Brands and publishers have limited control of the whole experience (as well as the branding) and they share the same space with competitors.

Blippar has been doing a very good job of providing high quality experiences on their browser. Operating like an agency, they take care of the end to end solution but publish on Blippar’s app. Layar is probably the king in terms of volume of experiences inside their platform. But with volume and scale, comes some limitations on the versatility of the experiences as you cannot go one-by-one. It remains to be seen if a new Layar-Blippar browser app will lean more towards volume or towards curated content.


So what lies in store in the future? Beautifully crafted, branded AR-capable apps are more likely to win hearts-and-minds of mobile users.A foreseeable option is that, just as happened with location, AR and image recognition capabilities will increasingly be embedded within a multitude of apps in a seamless fashion. (Credit to D.Marimon for parts of this post).


Wednesday, July 31, 2013

How can agencies produce great Augmented Reality campaigns?


If you are an agency (creative, advertising, mobile or digital) and you are new to augmented reality (AR) it is not easy to decide how to convert a creative idea into a fully fledged AR app.

Where do you start? What AR components do you need? What are the ingredients for a great AR app, be it a game, a marketing app or a functional app? What would it cost and how long will it take? How do you measure the Return on Investment (ROI) for the app?

Here is are 4 pointers that you may find helpful:

1. AR DOES NOT NEED TO BE COMPLICATED

Yes, it's true. Mention AR and instinctively people conjure images of 3D objects flying out of a mobile screen with complex graphic animation. True, you could do this but...it is not always essential. Linking physical objects like a packet of cereals to a TV advert can be done easily and economically simply by using image recognition or tracking (tracking will allow you to view the video on top of the box instead of opening a new screen on your device).

2. WHAT DO YOU NEED TO CREATE A GREAT AR APP?

Here is the tricky part -there are several components you could require for an AR app depending on the complexity of what you're looking to do. Take the IKEA AR shop catalogue for 2013 for example. This catalogue has 43 individual images dotted throughout the printed edition that link to digital content. Some link to videos (if you don't have these, you will need to produce them), some link to web pages (virtual stores where you can purchase items) and others still link to 3D animated computer graphics.

The production of the AR app required:

-a Tracking SDK
-an Image Recognition SDK
-creative computer graphics content
-mobile app development
-an AR creator platform to link "virtual buttons" to actual actions within the mobile app

3. WHAT SHOULD THE APP DO?

This links to the first point..."keep it simple" is my advice. It is clear some use cases work better than others and that this varies according to who you are targeting. AR for retail and print works great. Augmenting a magazine, newspaper or catalogue offers clear, tangible benefits to the reader -you see new content that is helpful or entertaining. AR for retail apps makes sense since indoors, GPS doesn't work, so using physical objects or shop signs as pointers for the app to perform an action is a clever workaround. Scan the poster displayed in the shop window and see special offers, for example.

4. WHAT SHOULD AN AR APP COST AND TAKE TO ROLLOUT?

It depends but....simple AR apps can be developed for around $20-25,000 as a very general rule of thumb. Complex ones can easily cost $100,000+ (hence the keep it simple rule!). A marketing app will offer an intangible return in terms of brand awareness and brand loyalty. Other AR apps generate revenue and more than pay for themselves in terms of ROI: a museum app that acts as a guide (audio and video) and lets you scan the popular exhibits to get AR content when you want it, is often offered at a premium in app stores and can have a short payback period. 

Wednesday, July 10, 2013

Augmented Reality 2.0 - From AR 1.0 (Hype) to AR 2.0 (Mainstream)

BARCELONA -I recently presented an overview of Augmented Reality to an interested customer describing some examples of note-worthy AR campaigns and how cloud-based image recognition helps solve many requirements of AR campaigns.

In fact, I believe we are now in AR 2.0, whereby AR is coming into the mainstream (thanks also to Google Glass) underpinned by powerful smartphones and 4G networks. Brands are coming on board, since consumers increasingly demand (and expect!) the physical products of their favourite brands to have a digital life. 

This AR experience can vary from a simple video or photo overlay to more complicated (3D) animations, though the latter have more of a gimmicky effect. Making AR work for brands means keeping it simple and focusing on the quality of the experience. This is where image recognition becomes key, being the gateway technology for AR and opening the path to enhance the user´s view of the world. 

It is also time to throw away notions of QR codes on print and objects being scanned by a user (nothing wrong however with users with QR codes on their phone being scanned themselves, e.g. for a airplane boarding pass). Applying QR codes for  marketing use cases never made much sense and was used for lack of a better option. Consumers everywhere (except Asia perhaps) have given a resounding no to this technology.

You can see below a selection of slides from my presentation: 

Friday, January 25, 2013

RIP QR Codes –why code-based scanning will give way to image recognition for mobile apps





Two quicks facts to kick things off: QR Codes (Quick Response Codes) were not designed for mobile advertising. They were created by the automotive industry to track vehicles as they flowed through the logistics chain.

While a significant advance at the time (1994) because they improved the scanning effectiveness (compared to barcodes) and could hold 100 times more data, today QR codes draw mixed reactions from consumers and developers alike.

According to Comscore’s survey of December 2011, only 20% of Americans use QR codes (this drops to 12% in Spain). The bottom line is that QR codes are not practical for average consumers to use and they are not the best way  for advertisers to connect with their customers.

Ever tried scanning a QR Code? You will need both a steady hand and some trial and error in finding the best angle for scanning. You will also need to find a Code Scanner app and download it to your phone, as well as to know how to use it! The barriers are significant. Not to mention the fact that QR codes can be hacked for malicious purposes, as happened in Russia.

The new way of recognizing images within adverts (in posters, magazine ads or on packaged goods) is by Image Recognition. Catchoom (based in Spain) is leading the way in this, using a Cloud-based recognition system to achieve image matching in under 0.5 seconds. The user simply points their smartphone’s camera at the image (for example, a logo), Catchoom processes the information with its patented image recognition algorithms, and hey presto! In less than a second, you get a weblink, an image or a video appearing on the phone screen. Compared to QR Codes, the technology really kicks ass, because what are known as “false positives” are vastly reduced. This means, around 99% of users pointing at an image to be recognized will get a perfect match and see the information the publisher has created.

The future has arrived...RIP QR codes.